When identifying roles and responsibilities for your new analytics personnel, it is important to start with the “why” in mind. Start with the customer. This could be a line of business within your organization or a government component. Because most data analytics teams are enabling a parent or larger organization, roles and responsibilities must be defined with a clear understanding of who the clients are and what business initiatives they are advancing. This will help categorize the broad application of analytics for which your organization will be responsible (i.e., marketing and sales, customer retention and personalization, supply chain and logistics, pricing and revenue management, etc.). It is also important to understand what other analytics organizations might already exist and how your organization will bring a unique perspective to the business need and client. Next, define how the new analytics team enables the client or business initiative. Most analytics teams will focus on:
- Building big data collection and analytics capabilities to uncover customer, product, and operational insights
- Analyzing data sources and proposing solutions to strategic planning problems on a one-time or periodic basis
- Providing data-driven decision support
- Developing analytics models and insights for customer- or employee-facing applications to drive efficiency or revenue
Finally, define what “types of analytics” the new analytics team will provide. Generally categorized along the spectrum of descriptive, predictive, and prescriptive analytics, your organization will need to define where along this data science spectrum the new team will operate. Where you land on this spectrum needs to be defined with consideration to the analytic maturity of your client/organization and what other analytics teams are already doing within the defined application. Also, consider the talent that your organization will have access to and how the talent falls within the analytic spectrum. If you're wondering how organizations can get started with advanced analytics, it's probably best to start with descriptive databases then move on to predictive models and fully prescriptive analytics. Much of the preparatory steps in building a basic analytics strategy will help as you move into dealing with increasingly complex business processes and analytics tools. For example, a business case that starts with a need to understand who your customers are (descriptive) could end up informing your approach to a business case that starts with a need to automatically recommend a product in real time that a consumer is likely to be interested in (prescriptive). Understanding customer behavior comes from first being able to identify the customer.
It is also important to recognize that analytics teams often depend on other enabling teams to move initiatives/projects along. One such enabling team is information technology and building a strong relationship with your IT organization (if external from your organization) will be critical to your success. Analytics leaders must collaborate with enabling teams when defining the analytic roles and responsibilities for your new analytics professionals. Consequently, enabling teams can successfully collaborate toward the greater vision or business initiative. Most importantly, analytics professionals, whether centralized or decentralized, should be working in partnership and maintaining organizational alignment on a per project basis with special emphasis on design thinking, collaboration, defining handoff procedures, and agreeing on how partner enabling teams account for value generation.
In a decentralized approach, analytics teams will also need to partner with other analytics teams to advance initiatives or projects; share talent, tools, and best practices; and build a community that encourages partnership and advances the vision to make your organization a data-driven organization. Other analytics teams might be a great way to augment capabilities if one analytics team is in need of expanding capabilities within the analytic spectrum or if you are resource-constrained. Citizen analytics teams (a Gartner-created term for individuals who have some rudimentary, “democratized” analysis skills but are not data scientists) might be good partners to help with adoption and implementation and can often be key players in project maintenance.
External vendors and university partnerships can also be instrumental. Building the talent and capability internally can be a long process, and external vendors and universities can provide the seed to get your team off the ground. In addition, these same partners can be instrumental when your analytics team is resource-constrained or working on expanding expertise into other applications, focus areas or moving along the analytic spectrum.